The 5D2 has strong pattern noise at ISO 100

tonylong wrote in post #8792471
But...but...well, we like to bicker!

yeah... we are becoming more like DPReview everyday

Jannie wrote in post #8792525
you have absolutely no idea what I can handle, but I seriously do hold back on what I can shove out

I have absolutely no idea, how such a technical topic can become a personal issue to someone. Are you a designer of the 5D2, or do you have Canon shares, or what else causes this?

I am asking this, because this attitude is not specific to you; it is quite common, I often see that some people can't stand a discussion they don't understand enough and even try to prevent others participating in that.

I don't think I've ever seen Canon advertise dynamic range in terms of stops, and "bit ADC" is not an expression of dynamic range.

kcbrown wrote in post #8784806
Heh. So clearly, if the shot were an interesting photo that couldn't have been obtained any other way but by taking a single shot, then everyone here would do an about face and agree that it's a problem.

Right....

:p

I'm usually much nicer than this, but seriously the only inidividual who has posted an example that illustrated this "problem" in a real, whorthwhile image was Ben. Anyone else care to do so?

Show me an image with interesting lighting that is worth posting, an image that one would be proud to post or print, and how this "problem" has ruined the image - then I'll see the point of this thread.

tonylong wrote in post #8784655
Daniel has noted that the problem hasn't been seen with the 7D. We don't know the root cause.

It's probably worth Daniel updating his first post I think as panopeeper recently highlighted a very similar issue with the 7D. Daniel has also contributed to that thread. That thread has also taken a similar route to this one unfortunately

versedmb wrote in post #8793079
Show me an image with interesting lighting that is worth posting, an image that one would be proud to post or print, and how this "problem" has ruined the image - then I'll see the point of this thread.

Most sensible thing said in this thread.

RDKirk wrote in post #8793017
I don't think I've ever seen Canon advertise dynamic range in terms of stops,

Neither have I, that I recall.

RDKirk wrote in post #8793017
...and "bit ADC" is not an expression of dynamic range.

By itself it is not, but the way Canon uses it is. This is what they say:

"14-bit A/D conversion (16,384 colors)"

This is what they are implying:

The camera will be capable of 16,384 colors.

This is what the truth is:

The 14-bit ADC is capable of 16,384 colors, but the camera itself has so much read noise that the actual number is less than 4096 (12-bit). Worse still, pattern noise will limit the "useful" bit range to even less.

The only way for the camera to be capable of 14-bit would be if its read noise was low enough.

versedmb wrote in post #8793079
I'm usually much nicer than this, but seriously the only inidividual who has posted an example that illustrated this "problem" in a real, whorthwhile image was Ben. Anyone else care to do so?

Show me an image with interesting lighting that is worth posting, an image that one would be proud to post or print, and how this "problem" has ruined the image - then I'll see the point of this thread.

How much evidence do you need to prove that this is a valid problem? 5 images? 10? 1000?

If you can prove a hypothesis is wrong in one situation, you've proven it wrong an infinite number of situations. I'm sure you've learned this in mathematics. The same idea applies here.

Having 10 stops of DR obviously isn't that important to you, so it's not a problem for you. And frankly, it's usually not a problem for me, either. That doesn't mean it's not a problem at all.

Daniel Browning wrote in post #8793399
This is what the truth is:

The 14-bit ADC is capable of 16,384 colors, but the camera itself has so much read noise that the actual number is less than 4096 (12-bit). Worse still, pattern noise will limit the "useful" bit range to even less.

The only way for the camera to be capable of 14-bit would be if its read noise was low enough.

But do they not compare iso 6400 on 5dm2 with iso3200 on 7d etc. What does this say about the 7d?

A question here is what Canon claims. Or what they try to imply.
http://www.usa.canon.c​om …ArticleAct&arti​cleID=2748

Note here that "information" in scientific language does not count noise.

When they introduced 14-bit conversion in the 1Ds3, they described it as:
http://www.usa.canon.c​om …egoryid=139&mod​elid=15710

This used-car salesman text, like there are 16,384 distinct brightness steps from the darkest to brightest tone (vs. 4,096 previously) is amusing. For example the 7D creates about 11560 "brightness steps" with ISO 100; this is apart from the noise aspect.

What's the difference between tonal range and dynamic range and what's the main emphasis with the 5d2, dynamic or tonal range?

Daniel Browning wrote in post #8793399
Neither have I, that I recall.

By itself it is not, but the way Canon uses it is. This is what they say:

"14-bit A/D conversion (16,384 colors)"

This is what they are implying:

The camera will be capable of 16,384 colors.

This is what the truth is:

The 14-bit ADC is capable of 16,384 colors, but the camera itself has so much read noise that the actual number is less than 4096 (12-bit). Worse still, pattern noise will limit the "useful" bit range to even less.

The only way for the camera to be capable of 14-bit would be if its read noise was low enough.

You are misunderstanding some vital parts here.

14 bit A/D converter doesn't mean 16.384 colors. Each color is made of three bytes, not one byte. If you have 14bit A/D converter, then you can have a color made of three 14 bit bytes. That is 16384^3, which is brutal number of colors.

The big lie is the quality of pixels when manufactures divide photocells in the sensors that much. The noise is obviously above the lowest measurements in such a sensors.

Nikolas wrote in post #8793990
What's the difference between tonal range and dynamic range and what's the main emphasis with the 5d2, dynamic or tonal range?

I think it is alike. Tonal range or digital dynamic range is build of power of two, the analogue DR is build of power 2.2 if I remember right, that is, by the power of logarithm.

5D_USER wrote in post #8794969
You are misunderstanding some vital parts here.

14 bit A/D converter doesn't mean 16.384 colors. Each color is made of three bytes, not one byte. If you have 14bit A/D converter, then you can have a color made of three 14 bit bytes. That is 16384^3, which is brutal number of colors.

The big lie is the quality of pixels when manufactures divide photocells in the sensors that much. The noise is obviously above the lowest measurements in such a sensors.

No, he is not misunderstanding any vital part.

If you find the discussion of a single color channel together with mentioning of 16.384 colors misleading, you should direct that complaint to Canon since it is a Canon quote you do not like.

They have a number of variations of the following on their web pages:

An educated reader would know that they should have said something like "16,384 shades per channel" but the error of talking about colors/channel is a Canon error.